Extended implied weighting

Several extensions to implied weighting, recently implemented in TNT, allow a better treatment of data sets combining morphological and molecular data sets, as well as those comprising large numbers of missing entries (e.g. palaeontological matrices, or combined matrices with some genes sequenced for few taxa). As there have been recent suggestions that molecular matrices may be better analysed using equal weights (rather than implied weighting), a simple way to apply implied weighting to only some characters (e.g. morphology), leaving other characters with a constant weight (e.g. molecules), is proposed. The new methods also allow weighting entire partitions according to their average homoplasy, giving each of the characters in the partition the same weight (this can be used for dynamically weighting, e.g. entire genes, or first, second, and third positions collectively). Such an approach is easily implemented in schemes like successive weighting, but in the case of implied weighting poses some particular problems. The approach has the peculiar implication that the inclusion of uninformative characters influences the results (by influencing the implied weights for the partitions). Last, the concern that characters with many missing entries may receive artificially inflated weights (because they necessarily display less homoplasy) can be solved by allowing the use of different weighting functions for different characters, in such a way that the cost of additional transformations decreases more rapidly for characters with more missing entries (thus effectively assuming that the unobserved entries are likely to also display some unobserved homoplasy). The conceptual and practical aspects of all these problems, as well as details of the implementation in TNT, are discussed.     

Weighting against homoplasy improves phylogenetic analysis of morphological data sets

The problem of character weighting in cladistic analysis is revisited. The finding that, in large molecular data sets, removal of third positions (with more homoplasy) decreases the number of well supported groups has been interpreted by some authors as indicating that weighting methods are unjustified. Two arguments against that interpretation are advanced. Characters that collectively determine few well‐supported groups may be highly reliable when taken individually (as shown by specific examples), so that inferring greater reliability for sets of characters that lead to an increase in jackknife frequencies may not always be warranted. But even if changes in jackknife frequencies can be used to infer reliability, we demonstrate that jackknife frequencies in large molecular data sets are actually improved when downweighting characters according to their homoplasy but using properly rescaled functions (instead of the very strong standard functions, or the extreme of inclusion/exclusion); this further weakens the argument that downweighting homoplastic characters is undesirable. Last, we show that downweighting characters according to their homoplasy (using standard homoplasy‐weighting methods) on 70 morphological data sets (with 50–170 taxa), produces clear increases in jackknife frequencies. The results obtained under homoplasy weighting also appear more stable than results under equal weights: adding either taxa or characters, when weighting against homoplasy, produced results more similar to original analyses (i.e., with larger numbers of groups that continue being supported after addition of taxa or characters), with similar or lower error rates (i.e., proportion of groups recovered that subsequently turn out to be incorrect). Therefore, the same argument that had been advanced against homoplasy weighting in the case of large molecular data sets is an argument in favor of such weighting in the case of morphological data sets. © The Willi Hennig Society 2008.     

"ALL TREE HISTOGRAMS" AND THE EVALUATION OF CLADISTIC EVIDENCE: SOME AMBIGUITIES

Abstract— Inspection of trees of varying lengths (by the option ALL TREES, which produces a histogram for tree lengths in PAUP 3.0) has been used to evaluate cladistic data and results. For example, a result may be judged more effective if the groups supported in the most parsimonious tree are preserved in trees that require increasingly greater amounts of homoplasy. Evaluation of grouping purely on the basis of this stability criterion ignores other highly relevant aspects of cladistic results. In particular, some data sets may incorporate additional taxa that introduce homoplasy to the shortest tree in a manner that concurrently allows for a revised understanding of character optimization patterns. These taxa may render groups preserved in the shortest tree less stable, but this result is not necessarily deficient if the homoplasy underlying such instability reveals possible character state changes for the given taxa irretrievable from the original matrix. The hypothetical example described here is relevant to so called "stem", "basal" or "plesiomorphic sister" taxa that are commonly considered in studies of both fossil and extant taxa.     

Soft anatomy, diffuse homoplasy, and the relationships of lizards and snakes

Most previous phylogenetic analyses of squamates (‘lizards’ and snakes) employing large character sets have focused on osteology. Soft anatomical traits bearing on this problem have usually been considered in small subsets. Here, a comprehensive phylogenetic analysis of squamate soft anatomy is attempted. 126 informative characters are assessed for 23 squamate lineages, representing snakes, amphisbaenians, dibamids, and all the traditionally recognized ‘families’ of lizards. The traditionally recognized groupings Iguania, Scleroglossa, Gekkota, Scincomorpha, Anguimorpha and Varanoidea are corroborated in this analysis. More controversial taxa are resolved as follows. Xantusiids, amphisbaenians and dibamids cluster with gekkotans, and snakes are strongly allied with anguimorphs in general, and varanids in particular. Nearly all these clades are congruent with those found in a recent comprehensive osteological analysis; the strong support for snake‐varanid relationships found in both studies is particularly notable. This congruence is surprising given that previous studies of soft anatomy tended to give differing and often heterodox results. These previous results can be attributed to overrepresentation of misleading characters in small isolated data sets. Such misleading signals are minimized when data sets are combined. For instance, the snake‐varanid clade is contradicted by many characters, and analyses of particular organ systems therefore give differing results. However, characters that are incongruent with the snake‐varanid clade also disagree with each other (diffuse homoplasy), rather than forming coherent support for some particular alternative clade (concerted homoplasy). In a combined analysis these incongruent but diffuse characters cancel each other out to leave a very strong (and orthodox) phylogenetic signal. These results underscore the view that the raw amount of homoplasy — as revealed by consistency and retention indices — is not the only determinant of phylogenetic signal; the distribution of that homoplasy is also important. Thus, questioning a phylogenetic hypothesis (e.g. the snake‐varanid clade) by identifying numerous conflicting characters is insufficient — the structure of the conflicting characters should be assessed in a rigorous phylogenetic analysis.     

IN SEARCH OF HOMOPLASTIC TENDENCIES: STATISTICAL INFERENCE OF TOPOLOGICAL PATTERNS IN HOMOPLASY

The “tendency” for homoplasy to appear in closely related taxa has been widely discussed but rarely quantified. This paper proposes statistical tests that examine the topological distribution of homoplasy within characters in phylogenies. They test whether character changes are localized (confined to some subtree), or clustered (occur in proximity to each other), relative to two null models of character evolution. Null Model I assumes that the observed number of character changes are dispersed randomly among the internodes of the tree, whereas Model II weights the probability that an internode contains a change by the length of that internode—estimated by the total number of character changes along that internode. Localization is measured by the largest furthest-neighbor distance between changes, clustering by the mean nearest neighbor distance. Distances are measured either by the number of intervening branches or the number of intervening character changes. Analyses of four cladistic data sets from the literature reveal very few characters that exhibit significant levels of clustering or localization—no more than would be expected by chance. In every data set a majority of characters exhibited at least weak tendencies, but in only one data set was there a significant excess of such characters. The present findings do not provide compelling evidence for the existence of “tendencies” in homoplasy, at least among characters used to reconstruct phylogenies. They should be sought elsewhere, in cladistic analyses of larger scope, probably among a class of characters defined a priori on a structural or functional basis.     

Alu insertion loci and platyrrhine primate phylogeny

Short INterspersed Elements (SINEs) make very useful phylogenetic markers because the integration of a particular element at a location in the genome is irreversible and of known polarity. These attributes make analysis of SINEs as phylogenetic characters an essentially homoplasy-free affair. Alu elements are primate-specific SINEs that make up a large portion of the human genome and are also widespread in other primates. Using a combination wet-bench and computational approach we recovered 190 Alu insertions, 183 of which are specific to the genomes of nine New World primates. We used these loci to investigate branching order and have produced a cladogram that supports a sister relationship between Atelidae (spider, woolly, and howler monkeys) and Cebidae (marmosets, tamarins, and owl monkeys) and then the joining of this two family clade to Pitheciidae (titi and saki monkeys). The data support these relationships with a homoplasy index of 0.00. In this study, we report one of the largest applications of SINE elements to phylogenetic analysis to date, and the results provide a robust molecular phylogeny for platyrrhine primates.     

The Phylogenetic Affinities of Nothofagus (Nothofagaceae) Leaf Fossils based on Combined Molecular and Morphological Data

The phylogenetic placements of leaf fossils of Nothofagus (Nothofagaceae) were determined using parsimony analyses of molecular and morphological data for extant species combined with morphological data for fossils. Placement was possible for only seven of the 30 or so described fossil species of Nothofagus because only these had sufficiently good preservation of both cuticular and leaf architectural characters. In combined analyses of morphology and molecular data, leaf cuticular characters showed little homoplasy. In contrast, many architectural characters, including some leaf margin and venation characters, showed high homoplasy, making it difficult or impossible to accurately determine the phylogenetic affinities of impression fossils of this genus.     

Phylogenetic relationships of spatangoid sea urchins (Echinoidea): taxon sampling density and congruence between morphological and molecular estimates

A phylogeny for 21 species of spatangoid sea urchins is constructed using data from three genes and results compared with morphology-based phylogenies derived for the same taxa and for a much larger sample of 88 Recent and fossil taxa. Different data sets and methods of analysis generate different phylogenetic hypotheses, although congruence tests show that all molecular approaches produce trees that are congruent with each other. By contrast, the trees generated from morphological data differ significantly according to taxon sampling density and only those with dense sampling (after a posteriori weighting) are congruent with molecular estimates. With limited taxon sampling, secondary reversals in deep-water taxa are interpreted as plesiomorphies, pulling them to a basal position. The addition of fossil taxa with their unique character combinations reveals hidden homoplasy and generates a phylogeny that is compatible with molecular estimates. As homoplasy levels were found to be broadly similar across different anatomical structures in the echinoid test, no one suite of morphological characters can be considered to provide more reliable phylogenetic information. Some traditional groupings are supported, including the grouping of Loveniidae, Brissidae and Spatangidae within the Micrasterina, but the Asterostomatidae is shown to be polyphyletic with members scattered amongst at least five different clades. As these are mostly deep-sea taxa, this finding implies multiple independent invasions into the deep sea.     

Taking phylogenetics beyond pattern analysis: can models of genome dynamics guide predictions about homoplasy in morphological and behavioral data sets?

Despite the considerable amount of interest in phylogeny reconstruction, patterns of homoplasy in morphological and behavioral data have received only limited attention to date, whereas the patterns of homoplasy in molecular data are relatively well understood. First, because the number of alternative molecular character states is strictly limited (particularly for nucleotide sequence data), higher rates of substitution generate higher levels of homoplasy. Second, depending on the relative proportions of constrained and unconstrained sites, each molecular data set has a time frame of applicability outside of which resolution becomes ambiguous. There is good evidence to suggest that numbers of alternative character states for morphological and even behavioral data may be similarly limited and that higher rates of evolution are often linked to higher rates of homoplasy. Like molecular data sets, morphological and behavioral data sets contain rapidly evolving characters as well as more conservative elements. Morphologies and behaviors related to sexual recognition and reproduction show low levels of intraspecific variation, but high levels of lability between species, making them crucial for species identification but often poor as markers of relationship at greater time depths. The organization theory of speciation derived by Carson is a model based on genome dynamics, and it predicts exactly this window of applicability for characters related to sexual reproduction. Nonsexual characters related to environmental adaptation should be applicable at greater phylogenetic depths. A better understanding of patterns of homoplasy enables a more sophisticated approach to the assessment of the relative reliabilities of alternative tree topologies.     

Sampling diverse characters improves phylogenies: Craniodental and postcranial characters of vertebrates often imply different trees

Morphological cladograms of vertebrates are often inferred from greater numbers of characters describing the skull and teeth than from postcranial characters. This is either because the skull is believed to yield characters with a stronger phylogenetic signal (i.e., contain less homoplasy), because morphological variation therein is more readily atomized, or because craniodental material is more widely available (particularly in the palaeontological case). An analysis of 85 vertebrate datasets published between 2000 and 2013 confirms that craniodental characters are significantly more numerous than postcranial characters, but finds no evidence that levels of homoplasy differ in the two partitions. However, a new partition test, based on tree‐to‐tree distances (as measured by the Robinson Foulds metric) rather than tree length, reveals that relationships inferred from the partitions are significantly different about one time in three, much more often than expected. Such differences may reflect divergent selective pressures in different body regions, resulting in different localized patterns of homoplasy. Most systematists attempt to sample characters broadly across body regions, but this is not always possible. We conclude that trees inferred largely from either craniodental or postcranial characters in isolation may differ significantly from those that would result from a more holistic approach. We urge the latter.     

Some applications of the LeQuesne compatibility test

The treatment advocated requires coding of every multistate character in additive binary form and accepts, but does not require, assignment of direction to the transformations. For each binary character is computed the ratio of the number of incompatibilities observed to the number of incompatibilities expected on the null hypothesis of the random distribution of its states. At this stage the 'noisy' characters are identified by their high ratios. These ratios are used as the basis for selecting a set of core characters common to several large compatible sets (cliques), plus the several characters particular to each large clique, followed by those characters which add further resolution or corroboration with low levels of homoplasy. The 'labelling' procedure of Guise, Peacock & Cleaves (1982) is adapted to identify 'noisy' taxa and individual character scores which are likely to be homoplastic. The procedure facilitates recognition of the robust portions of a dendrogram, alternative further resolutions, and remaining areas of uncertainty.
For the purposes of illustration, data for African toads of the Nectophrynoides lineage are analysed and the resulting dendrogram is plotted on the map of Africa. Some observations are also made on data for lizards of the family Pygopodidae and on birds of the family Alcidae.     

Levels of Homoplasy in the Evolution of the Mammalian Skeleton

It is commonly believed that there are differences in the evolutionary lability of the crania, dentition, and postcrania of mammals, the latter two being more prone to homoplasy because of strong selective pressures for feeding and locomotion, respectively. Further, because of the fragmentary nature of fossils, phylogenetic analyses of extinct taxa often must utilize characters based on only one of these systems. In this paper the levels of homoplasy (as measured by the consistency index; CI) were compared in characters based on these three anatomical systems in therian mammals. No statistically significant differences were found in the overall CIs of 41 data sets based on dental, cranial, or postcranial characters. Differences in homoplasy within data sets with two or three kinds of data were not statistically significant. These findings suggest that dental, cranial, and postcranial characters can be equally prone to homoplasy and none should be automatically dismissed, disregarded, or systematically weighted in phylogenetic analyses. The level of homoplasy in characters derived from a given region of the skeleton may differ depending on the taxonomic level of the taxa considered. Dental, cranial, and postcranial characters may not constitute natural classes, yet examination of the phylogenetic signal of these subsets of data previous to a simultaneous analysis can shed light on significant aspects of the evolutionary process.     

PATTERNS OF VARIATION IN LEVELS OF HOMOPLASY

Patterns of variation in levels of homoplasy were explored through statistical analyses of standardized consistency indexes. Data were obtained from 60 recent cladistic analyses of a wide variety of organisms based on several different kinds of characters. Consistency index is highly correlated with the number of taxa included in an analysis, with homoplasy increasing as the number of taxa increases. This observation is compatible with a simple model of character evolution in which 1) the probability of character-state change increases with the total number of branches in a tree and 2) the number of possible states of a character is limited. Consistency index does not show a significant relationship to the number of characters utilized in an analysis or to the taxonomic rank of the terminal taxa. When the relationship between consistency index and number of taxa is taken into account, there is no significant difference between plant and animal data sets in the amount of homoplasy. Likewise, the level of homoplasy in morphological and molecular data sets does not appear to differ significantly, although there are still too few molecular studies to be confident of this result. Future comparisons of consistency indexes, including studies along the lines established here, must take into account the influence of the number of taxa on homoplasy.     

Morphological innovation through gene regulation: an example from Devonian Onychodontiform fish

The recent development of novel phenotypic designs by changes in gene regulation has been extensively discussed within the context of evolution and development. The fossil record shows many new designs (or body plans) which appear rapidly at various stratigraphic levels. One example of this is the arrival of the lobe finned fish in the Early Devonian, when a great variety of new forms appeared. These include the dipnoans, the onychodontids, the porolepiforms and the osteolepiforms, which differ widely in a number of characters. Each of these groups originated from an unspecified sarcopterygian source and they have since evolved independently. They also carry over the primitive genes of the parent or parents and similar changes in these genes will not produce synapomorphies between members of the different lineages. Unless care is exercised, homoplasies will be used as synapomorphies. Evidence must be found to find groups of features that define uniform functional entities. These define monophyletic groups. Recognition of homoplasy of characters thus becomes important. The study of the new functional structures and the areas from which they were derived by changes in gene regulation, would give us more evolutionary information.     

Exhaustion of morphologic character states among fossil taxa

Frequencies of new character state derivations are analyzed for 56 fossil taxa. The hypothesis that new character states are added continuously throughout clade history can be rejected for 48 of these clades. Two alternative explanations are considered: finite states and ordered states. The former hypothesizes a limited number of states available to each character and is tested using rarefaction equations. The latter hypothesizes that there are limited possible descendant morphologies for any state, even if the character has infinite potential states. This is tested using power functions. The finite states hypothesis explains states: steps relationships significantly better than does the ordered states hypothesis in 14 cases; the converse is true for 14 other cases. Under either hypothesis, trilobite clades show appreciably more homoplasty after the same numbers of steps than do molluscs, echinoderms, or vertebrates. The prevalence of the exhaustion pattern among different taxonomic groups implies that worker biases are not to blame and instead implicates biological explanations such as intrinsic constraints or persistent selective trends. Regardless of the source of increased homoplasy, clades appear to exhaust their available character spaces. Nearly all examined taxa show significant increases in proportions of incompatible character pairs (i.e., those necessarily implying homoplasy) as progressively younger taxa are added to character matrices. Thus, a deterioration of hierarchical structure accompanies character state exhaustion. Exhaustion has several implications: (1) the basic premise of cladistic analyses (i.e., that maximum congruence reflects homology rather than homoplasy) becomes increasingly less sound as clades age; (2) sampling high proportions of taxa probably is needed for congruence to discern homoplasy from homology; (3) stratigraphic data might be necessary to discern congruent homoplasy from congruent homology; and (4) in many cases, character states appear to have evolved in ordered patterns.     

The relative lability of floral vs. non-floral characters and a morphological phylogenetic analysis of Cobaea (Polemoniaceae)

The relative levels of lability in floral vs. vegetative characters have been suggested to give insight into the mechanisms of adaptation and speciation. Cobaea (Polemoniaceae) exhibits a remarkable diversity in floral form. A morphology-based phylogeny was constructed and is congruent in most regards with a previously published molecular phylogeny, but significant incongruence was found in the placement of two taxa. A combined analysis was performed, excluding C. aquatoriensis and C. lutea, and was nearly identical to the molecular analysis. The phylogenies are compatible with a recently published classification. Incongruencies between the phylogenies within section Rosenbergia may have profound implications for floral evolution. A simple method for testing the levels of homoplasy, as an indicator of lability between two classes of data, is developed and used to test for differences between floral and non-floral classes of characters. The method is based on repeated randomization of the data into partitions and recalculation of a test statistic defined as the difference between consistency index (CI) values of the data in each partition over a given tree. The null hypothesis that the floral characters are no more homoplasious than the non-floral characters cannot be rejected (P=0.09).     

THE USEFULNESS OF BEHAVIOR FOR PHYLOGENY ESTIMATION: LEVELS OF HOMOPLASY IN BEHAVIORAL AND MORPHOLOGICAL CHARACTERS

It is widely believed that behavior is more evolutionarily labile and/or more difficult to characterize than morphology, and thus that behavioral characters are not as useful as morphological characters for estimating phylogenetic relationships. To examine the relative utility of behavior and morphology for estimating phylogeny, we compared levels of homoplasy for morphological and behavioral characters that have been used in systematic studies. In an analysis of 22 data sets that contained both morphological and behavioral characters we found no significant difference between mean consistency indices (CIs, which measure homoplasy) within data sets for the two types of characters. In a second analysis we compared overall CIs for 8 data sets comprised entirely of behavioral characters with overall CIs for 32 morphological data sets and found no significant difference between the two types of data sets. For both analyses, 95% confidence limits on the difference between the two types of characters indicate that, even if given the benefit of the doubt, morphological characters could not have substantially higher mean CIs than behavioral characters. These results do not support the idea that behavioral characters are less useful than morphological characters for the estimation of phylogeny.     

Morphological phylogenetics of the sea spiders (Arthropoda: Pycnogonida)

Pycnogonids or sea spiders are a group of marine arthropods whose relations to the chelicerates have been an issue of controversy. Higher-level phylogenetic relationships among the lineages of sea spiders are investigated using 36 morphological characters from 37 species from all extant families and a Devonian pycnogonid fossil. This is one of the first attempts to analyze the higher-level relationships of the Pycnogonida using cladistic techniques. Character homoplasy (implied weights) is taken into account to construct a polytomous, most-parsimonious tree in which two major clades within Pycnogonida are obtained. Clade A includes Ammotheidae paraphyletic with Colossendeidae, Austrodecidae and Rhynchothoracidae, and clade B is formed by Nymphonidae, Callipallenidae (apparently paraphyletic), Pycnogonidae and Phoxichilidiidae. The analysis of equally weighted data is presented and helps to identify those characters less consistent. The reduction of the chelifores, palps and ovigers — shown independently within each of the clades as parallel evolution events — challenges the assumption of a gradual mode of reduction within the group, according to analysis of unordered vs ordered characters. Most of the phylogenetic affinities proposed here are compatible with traditional classifications. However, traditional taxonomic characters need to be complemented by sets of anatomical, molecular and developmental data, among others, to produce more robust phylogenetic hypotheses on the higher- and lower-level relationships of the sea spiders.     

Higher-Order Homoplasy Tests

The Le Quesne test of character compatibility uses pairwise comparisons of characters to detect homoplasy in phylogenetic character data. If a pair of characters fails this test we can conclude that a minimum of a single extra step is required by the pair of characters. The rationale of the Le Quesne test is extended to comparisons of triplets of characters. The triplet homoplasy test can reveal that that there is a minimum of four extra steps across a triplet of characters and thus that there are at least two extra steps associated with one of the characters. The triplet homoplasy test can thus detect higher orders of homoplasy than can be detected by the pairwise Le Quesne test. The possibility of quartet and other higher-order homoplasy tests is discussed. The utility of higher-order homoplasy tests is discussed. It is suggested higher-order homoplasy tests have potential uses analogous to the uses of the Le Quesne test, particularly with respect to data exploration.     

Reconstructing plumage evolution in orioles (Icterus): repeated convergence and reversal in patterns

Several empirical studies suggest that sexually selected characters, including bird plumage, may evolve rapidly and show high levels of convergence and other forms of homoplasy. However, the processes that might generate such convergence have not been explored theoretically. Furthermore, no studies have rigorously addressed this issue using a robust phylogeny and a large number of signal characters. We scored the appearance of 44 adult male plumage characters that varied across New World orioles (Icterus). We mapped the plumage characters onto a molecular phylogeny based on two mitochondrial genes. Reconstructing the evolution of these characters revealed evidence of convergence or reversal in 42 of the 44 plumage characters. No plumage character states are restricted to any groups of species higher than superspecies in the oriole phylogeny. The high frequency of convergence and reversal is reflected in the low overall retention index (RI = 0.66) and the low overall consistency index (CI = 0.28). We found similar results when we mapped plumage changes onto a total evidence tree. Our findings reveal that plumage patterns and colors are highly labile between species of orioles, but highly conserved within the oriole genus. Furthermore, there are at least two overall plumage types that have convergently evolved repeatedly in the three oriole clades. This overall convergence leads to significant conflict between the molecular and plumage data. It is not clear what evolutionary processes lead to this homoplasy in individual characters or convergence in overall pattern. However, evolutionary constraints such as developmental limitations and genetic correlations between characters are likely to play a role. Our results are consistent with the belief that avian plumage and other sexually selected characters may evolve rapidly and may exhibit high homoplasy. The overall convergence in oriole plumage patterns is an interesting evolutionary phenomenon, but it cautions against heavy reliance on plumage characters for constructing phylogenies.